ThisCalculator

Everyday

Active Backup for AI: Trustworthy Data Guide

Updated · Standard formula, shown below

Active Backup for AI: Trustworthy Data Guide

This guide explains how to configure Active Backup for Business (or similar backup tools) to make your AI training data more trustworthy. You’ll learn how to set up versioned, verified backups that protect against data corruption, accidental deletion, and ransomware—ensuring your AI models are trained on reliable, consistent data. We’ll cover the essential steps, common pitfalls, and troubleshooting tips.

What You’ll Need

  • Active Backup for Business installed on a Synology NAS (or comparable backup software on your server).
  • A source device (PC, server, or file server) containing the data you want to back up.
  • Sufficient storage space on the backup destination (NAS or cloud).
  • Administrator access to the backup software and the source device.
  • Basic understanding of your data’s structure and where it lives.

Step-by-Step

  1. Identify your critical data – Determine which datasets are essential for AI training. This could be raw training data, cleaned datasets, or model checkpoints. List their locations and sizes.
  2. Create a dedicated backup task – In Active Backup for Business, go to the “File Server” or “PC” tab (depending on your source) and click “Create Task.” Choose a name that clearly identifies the data, e.g., “AI_Training_Data_v1.”
  3. Select the source and destination – Specify the source folder(s) containing your AI data. For the destination, choose a local shared folder on the NAS or a connected cloud storage. Ensure the destination has enough free space.
  4. Enable versioning – In the task settings, enable “Versioning” or “Keep multiple versions.” Set a retention policy that suits your needs—for example, keep daily versions for 30 days and monthly versions for 12 months. This allows you to roll back to a previous state if data becomes corrupted.
  5. Set a backup schedule – Configure the backup to run regularly. For AI data that changes frequently, consider daily or even hourly backups. For static datasets, weekly might suffice. Consistency is key to ensuring you always have the latest trustworthy snapshot.
  6. Enable integrity checking – Turn on “Data Integrity Check” or “Checksum Verification” in the backup settings. This ensures that the backed-up files are not corrupted during transfer or storage. Active Backup for Business does this automatically for some tasks, but verify it’s enabled.
  7. Add a ransomware protection plan – If your NAS supports it, enable “Snapshot Replication” or “Immutable Backups.” This creates point-in-time snapshots that cannot be modified or deleted by ransomware, ensuring your AI data remains safe even if the primary source is compromised.
  8. Run an initial backup – Execute the first backup manually to ensure everything works. Monitor the process for errors. Once completed, review the backup log to confirm all files were transferred successfully.
  9. Test a restore – Regularly perform a test restore to a separate location. This verifies that your backups are not only present but also usable. A common mistake is assuming backups work without ever testing them.
  10. Document your backup and restore procedures – Write down the backup configuration, retention policies, and restore steps. This ensures that anyone on your team can recover data if needed, and it helps maintain trust in your AI pipeline.

Common Mistakes

  • Not enabling versioning – Without versioning, a single corrupted file overwrites the good one, and you have no way to roll back.
  • Backing up to the same disk as the source – If the disk fails, both the original and backup are lost. Always use a separate physical drive or cloud.
  • Ignoring backup errors – Many users see a warning but ignore it. Always investigate and resolve errors immediately.
  • Forgetting to test restores – Backups that are never restored are just guesses. Test at least quarterly.
  • Not excluding temporary or cache files – Including unnecessary files bloats the backup and can introduce inconsistencies. Exclude system and temp folders.
  • Overlooking file permissions – If the backup service lacks read permissions on some files, those files are silently skipped. Check the logs for permission warnings.

Tips

  • Use a naming convention for versions – Include dates or version numbers in your backup task names to make identification easy.
  • Monitor backup health – Set up email notifications for backup failures. This way, you know immediately if something goes wrong.
  • Leverage incremental backups – Active Backup for Business uses incremental backups by default, which saves space and time. Ensure this is enabled.
  • Store a copy offsite – For maximum trustworthiness, replicate your backups to a different location (e.g., another NAS or cloud service). This protects against site-wide disasters.
  • Keep an eye on storage usage – Versioning can consume a lot of space. Regularly review your retention policy and adjust if needed.
  • Integrate with your AI pipeline – Automate the backup to run before any major data processing step, so you always have a clean pre-processing snapshot.

Troubleshooting

Backup fails with “Access Denied” – Check that the backup service account has read permissions on the source folder. On Windows, you may need to run the backup service as a user with sufficient privileges. On Linux, verify the user has read access to the directory and files.

Backup is slower than expected – If your initial backup is large, it may take a while. Ensure the source and destination are on fast connections. For network transfers, consider using a wired connection or increasing bandwidth limits in the settings.

Versioning is not working – Verify that versioning is enabled in the task settings. Also, check that the destination file system supports versioning (e.g., Btrfs on Synology). If you’re using an external drive with a different file system, versioning may not be supported.

Restore fails or produces corrupted files – This usually indicates a problem with the backup itself. Run a data integrity check on the backup task. If the check fails, you may need to delete the backup and run a fresh full backup.

Backup task disappears or won’t start – Restart the Active Backup for Business service from the NAS control panel. If the task is still missing, recreate it. Also, ensure the source device is online and reachable.

Frequently Asked Questions

Does Active Backup for Business support versioning?

Yes, it does. You can enable versioning in the task settings, and it uses Btrfs snapshots on compatible NAS models to keep multiple versions efficiently.

Can I back up to a cloud service?

Yes, Active Backup for Business supports backing up to Synology C2 Storage, and you can also use Hyper Backup to replicate to other clouds like AWS S3 or Google Cloud.

How often should I run backups for AI training data?

It depends on how frequently your data changes. For active projects, daily backups are recommended. For static datasets, weekly is sufficient. The key is to have a recent, reliable snapshot.

What is the difference between a backup and a snapshot?

A backup is a copy of your data that can be stored separately. A snapshot is a point-in-time capture of the file system state, usually on the same storage. Snapshots are useful for quick recovery, but backups protect against hardware failure.

How do I ensure my backups are not corrupted?

Enable the built-in data integrity check (checksum verification) in your backup task. Additionally, periodically test restores to confirm the data is usable.

Can I schedule backups to run automatically?

Yes, you can set a schedule in the task settings. You can choose daily, weekly, or custom intervals, and even run multiple backups per day if needed.

Related

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.